Evaluation and Certification

The rulemaking process launches a model; evaluation decides its fate. And for the rare model that succeeds, the law provides an unusual reward: national expansion without an act of Congress, but only if it clears a specific, rigorous bar. This lesson connects the policy process to the evaluation course’s standard.

Every model is evaluated

Section 1115A requires models to be evaluated, and CMMI contracts independent evaluations that apply exactly the methods the evaluation course covered: comparison groups, difference-in-differences, and honest accounting of gross versus net savings. Those evaluations, plus the CBO’s budget scoring, are the evidence base on which a model’s future is decided.

The certification bar

The pivotal moment for a successful model is certification for expansion:

Worth remembering: the law lets the Secretary expand a model nationally only if the CMS Chief Actuary certifies that expansion would reduce net spending, or at least not increase it, without harming quality, and the Secretary finds it would not deny or limit coverage. This is the “did it work” question from the evaluation course turned into a legal standard, and it is deliberately hard. It requires net savings, verified independently, not a favorable pilot or an encouraging press release.

Why so few clear it

Of roughly 70 models tested, only about four have been certified for national expansion. That scarcity is not obstruction; it is the evaluation reality of the whole curriculum made official. When you apply a rigorous, independently verified, net-savings standard, most models do not meet it, for all the reasons the evaluation course laid out: selection, regression to the mean, benchmark assumptions, and the gap between gross and net. And the CBO found that CMMI as a whole increased net federal spending in its first decade. Certification is rare because real, net, verified savings are rare.

What this means for a participant

Certification is a powerful signal to read. A certified model has cleared the highest evidentiary bar in the program; an uncertified one, however promising its early results, has not. For anyone tracking or engaging with models, the certification status is a quick, meaningful filter: it separates the handful of models proven to save money net from the many that produced encouraging but unconfirmed results. That is the honest, evidence-first way to read the field, and the reason this course pairs the policy process with the evaluation discipline.

Key takeaways

• Section 1115A requires model evaluation, using the comparison-group and net-savings methods from the evaluation course.
• National expansion requires an independent actuarial certification of net savings without quality harm, a deliberately high bar.
• Only about four of roughly seventy models have been certified, reflecting how rare real, verified net savings are.

Sources

• Congressional Budget Office, Federal Budgetary Effects of the Activities of the Center for Medicare and Medicaid Innovation (September 2023) (opens in a new tab)
• Social Security Act, Section 1115A (opens in a new tab)